A multiple-case study using a qualitative approach was used to explore how elementary age students used voice-to-text (i.e., Google Voice Typing) to compose.
How students responded to using voice-to-text as a method of composing, the effect of voice-to-text on features (i.e., content, appearance, etc.) of composition, and the change of the composing process for students was examined.
This chapter provides a rationale for design, the selection process, the data collection methods, and data analysis. In addition, the chapter summarizes the strategies used to ensure trustworthiness.
Theoretical Framework
Epistemology is what one believes about the nature of knowledge, that is, the belief of what knowledge is and how it is created. I adopt constructionism as my epistemological stance because I believe reality is socially constructed. Crotty (2015) states, constructionism
…is the view that all knowledge, and therefore all meaningful reality as such, is contingent upon human practices being constructed in and out of interaction between human beings and the world, and developed and transmitted within an essentially social context. (p. 42).
Meaning is constructed as humans interact and engage with their surroundings. The world and objects in the world have no meaning until this happens. Once the human mind interacts with the world and objects in the world, meaning is constructed. In other words, through consciousness, objects acquire meaning.
Ontology is what one believes about the nature of reality. It is my belief that meaning is not constructed in one single way, but “…there are multiple realities, or interpretations, of a single event” (Merriam & Tisdell, 2016, p. 9). As the researcher, I come into the study with my own set of realities, as does each participant and each reader. Therefore, as Creswell (2013) suggests, “the use of multiple forms of evidence in themes using the actual words of different individuals and presenting different perspectives” (p. 20) are included in this study.
My epistemological stance influences the theoretical framework for this study.
Because I am interested in describing, understanding, and interpreting the experience of a phenomenon from the perspective of an individual, I take an interpretive or
constructivist perspective. According to Crotty (2015), “…constructivism is primarily an individualistic understanding of the constructionist position…” (p. 58). He goes on to say, “Constructivism describes the individual human subject engaging with the objects in the world and making sense of them” (p. 79). That is, constructivism is an individual’s way of interacting and understanding the world. Every individual
constructs knowledge in a unique way, unlike anyone else. “Each reality is somewhat different, because each person’s experiences and resulting apperceptions are different”
(Jonassen, 1991, p.7).
In much the same way, “…writers don’t find meaning, they make [meaning]”
(Flower & Hayes, 1980, p. 21). Writing is constructing ideas, not finding them.
Skilled writers construct their own unique representation of the rhetorical problem (i.e., assignment, audience, goals), meaning they understand what they are trying to do as they write. “If we can teach students to explore and define their own problems, even
within the constraints of an assignment, we can help them to create inspiration instead of wait for it” (Flower & Hayes, 1980, p. 32). Romano (1987) stated, “…students write to discover, create and explore their thinking, dig up prior knowledge, to cultivate intellectual independence, to conjecture about possibilities, to struggle with difficult concepts, and to engage the imagination as an ally in learning” (p. 34). Finally, Sondra Perl said it so well,
Composing always involves some measure of both construction and discovery.
Writers construct their discourse inasmuch as they begin with a sense of what they want to write. This sense, as long as it remains implicit, is not equivalent to the explicit form it gives rise to. Thus, a process of constructing meaning is required. Rereading or backward movements become a way of assessing whether or not the words on the page adequately capture the original sense intended. Constructing simultaneously affords discovery. Writers know more fully what they mean only after having written it. In this way the explicit
written form serves as a window on the implicit sense with which one began. (as cited in Perl, 1994, p. 55)
Research Design
My research methodology is shaped by my epistemological stance and
theoretical framework. Qualitative researchers believe reality is socially constructed, not “out there” as one single, observable reality. The researcher’s goal is not to find knowledge but to construct it. There are four main characteristics as suggested by Merriam & Tisdell (2016):
• The researcher seeks to understand how people interpret their experience.
• The researcher is the primary instrument for data collection.
• Research is inductive, not deductive.
• The product is descriptive, not numeric.
Merriam & Tisdell (2016) state, “…qualitative researchers are interested in understanding the meaning people have constructed, that is, how people make sense of their world and the experiences they have in the world” (p. 15). Thus, the participant’s perspective and understanding the participant’s experience and what the experience was like for them is of importance. The focus of this study was based on each participant’s perspective of using Google Voice Typing to compose.
In qualitative research, the researcher is the primary instrument for gathering data. There are advantages and disadvantages to being the primary data collector.
Advantages include being able to “be immediately responsive and adaptive” and to
“expand…understanding through nonverbal as well as verbal communication, process the information (data) immediately, clarify and summarize material, check with respondents for accuracy of interpretation, and explore unusual or unanticipated responses” (Merriam & Tisdell, 2016, p.16). Yet, being so close to the research, the researcher’s biases can greatly impact a study. As the primary data collector, the researcher has no one but herself to rely on to make all the right decisions and
interpretations. For this reason, the researcher identifies and monitors any biases that arise during the study.
Qualitative research is inductive, not deductive, meaning data is gathered to build a hypothesis, not to test a hypothesis. Specific raw data from interviews, observations, or documents grow into abstract or general themes. It is important to
note, when testing a hypothesis, it is possible to predict the behavior; this is not the case for qualitative research, as variables are unknown ahead of time.
Lastly, the final product of qualitative research is descriptive, not numeric. The researcher communicates the findings through words or pictures, by describing the context, participants, or activities (Merriam & Tisdell, 2016). The final product uses quotes from the participant(s) or other specific information from the field notes and interview(s).
This qualitative study uses a multiple-case study design. A multiple-case study provides an in-depth description and analysis of several individual cases. There are two stages of analysis. First, a with-in case analysis gathers findings from individual case studies separately. Then, a cross-case analysis looks at all the cases together; building a general explanation that addresses each research question. A multiple-case design can strengthen the validity and stability of the findings (Merriam & Tisdell, 2016).
According to Merriam & Tisdell (2016), “A case study is an in-depth description and analysis of a bounded system” (p. 37). To check for boundedness, Merriam & Tisdell (2016) suggest asking two questions: (1) Is there a limit to the number of people involved? (2) Is there a finite time for observations? If the answer is
“yes” to both questions, the case is bounded. The bounded system, or unit of analysis, for this case study consisted of students in second to fourth grade, from a private, urban elementary school in a large southwestern city, who were either reluctant writers or had poor writing or handwriting skills. The study was bounded by time (2 months) and used multiple cases (four students attending a Technology Writing Workshop on separate occasions).
Participants
“Purposeful Sampling is based on the assumption that the investigator wants to discover, understand, and gain insight and therefore must select a sample from which the most can be learned” (Merriam & Tisdell, 2016, p. 96). To qualify for this study, participants had to be a second to fifth-grade student from a private urban elementary school in a large southwestern city, meeting one of the following additional criteria:
• reluctant writer (i.e., can write, but does not have the motivation to write);
• poor writing skills (e.g., disorganized or limited ideas, poor leads, plagiarism, unfocused writing, weak sentence structure, poor word choice, does the bare minimum);
• poor handwriting skills.
The school and principal gave permission for this study to be conducted. To recruit participants, an email was sent to second to fifth-grade teachers at the school.
The email explained the study as part of a Technology Writing Workshop that included instruction in writing and using Google Voice Typing. Additionally, the email asked teachers to identify students to participant in the study based on the above criteria and to send home a flier with students who met the criteria, encouraging interested participants to contact me. Three parents (one parent had two children) contacted me in person or by email about having their child participate in the study. The four participants selected for participation were Peniqua (3rd grader), Johny (2nd grader), Trixie (4th grader), and Sam (4th grader) (Table 1).
Table 1
Participant Information
Student Gender Grade Handedness Reluctant writer
Equipment. Participants used a MacBook Pro (13-inch, Late 2011) laptop (with 4GB of RAM and an Intel Core i5 processor, running at 2.4 GHz). To create compositions, participants used Google Docs, Google Voice Typing, a Blue Snowball USB Microphone or Apple EarPod Headphones. Google Voice Typing is a free continuous-speech dictation program built into Google Docs and only available in Chrome browsers.
Technology writing workshop. Parents of participants were sent a letter (Appendix A) with a description of the study, including the ethical guidelines for participation in the study. The letter explained (1) the purpose of the research; (2) what was expected of the participant; (3) the length of the study; (4) the risks and benefits;
(5) compensation for participating; and (6) protection of participant’s information and identity. The letter assured that participants could withdraw at any time and would not be penalized. Once parental permission was granted for participants to attend the Technology Writing Workshop, dates were set. Parents also gave permission for
participants to be audio and video recorded. Participants could withdraw at any time.
Participants attended the Technology Writing Workshop independently. The workshop was conducted at the participants’ school over five sessions. The procedural protocol below was followed for each participant to ensure procedural integrity and consistency (Table 2).
Table 2
Summary of Each Session
Session Time Setting Method Activity
1 about 2 Note. Handwritten, spontaneous = HWS; Handwritten, narrative Guide = HWNG;
Google Voice Typing = GV; Google Voice Typing, spontaneous = GVS; Google Voice Typing, narrative Guide = GVNG
Session One: Signed Child Assent (7-11 years), Attitude Survey, and Writing via Handwriting. Session one took place in my classroom, lasting about two
hours. To begin the session, the assent form (Appendix B) was read aloud to each participant, ensuring that each child understood his/her part in the study. I answered questions and had the child sign the form. After the participant signed the assent form and, in order to keep the participants’ identity anonymous and confidential, each participant chose a pseudonym to use in place of their real name.
To begin the Technology Writing Workshop, participants took the Elementary Writing Attitude Survey (Kear, Coffman, McKenna, Ambrosio, 2000). Next,
participants were introduced to the narrative genre by examining the elements of story structure (i.e., plot, setting, characters, etc.) in mentor texts. Participants discussed the story elements of each mentor text, analyzing how the author applied each element to the story. Then, participants were instructed to compose a spontaneous narrative writing. To honor the time given for each activity, a timer was set for each writing process. At the end of each time period, if participants needed more time, they were granted additional time. 15 minutes were given for prewriting and drafting. When time was up, participants had about 10 minutes to revise and edit their handwritten
compositions. At the end of 10 minutes, participants assessed their compositions, using a guide (Appendix C).
The second composition was a planned writing in which participants used a graphic organizer (Appendix D) to plan their handwritten narrative. I modeled and instructed participants on how to use the graphic organizer to plan the characters, plot, setting, theme, and point of view of their story. Participants had about 10 minutes to plan their narrative using the graphic organizer. After participants planned their
narrative, they had about 15 minutes to draft. When they were finished, they had about
10 minutes to revise and edit. Upon completion, participants assessed their composition using a guide (Appendix C).
Session Two: Training of Google Voice Typing. Session Two took place in my classroom, lasting about two hours. There were three stages of training. The first stage of training was learning about Google Voice Typing and how to use it. The process of dictating a composition using Google Voice Typing was modeled for each participant. During the demonstration I used a Google Voice Typing tutorial (Fudickar, 2017) and modeled how to start a new document, get the microphone, compose orally, and speak naturally but clearly with consistent volume and speed. I also demonstrated how to read aloud a story while dictating with Google Voice Typing. During the second stage of training, participants completed the same Google Voice Typing tutorial that was modeled for them. For the last stage of training, participants dictated a short story using Google Voice Typing. The purpose of dictating a story was to give
participants extra practice speaking clearly, dictating punctuation, and using commands.
As they dictated, they corrected mistakes with voice commands or by typing.
Session Three: Writing via Google Voice Typing. Session Three took place in my classroom, lasting about two hours. Once training was complete, participants began the second writing session. During this time, participants dictated two
compositions using Google Voice Typing. First, participants were instructed to dictate a spontaneous narrative writing. To honor the time given for each activity, a timer was set for each writing process. At the end of each time period, if participants needed more time, they were granted additional time. 15 minutes were given for prewriting and drafting. When time was up, participants had about 10 minutes to revise and edit their
handwritten compositions. At the end of 10 minutes, participants assessed their compositions, using a guide (Appendix C).
The second composition was a planned writing in which participants used a graphic organizer (Appendix D) to plan their narrative. I modeled and instructed participants on how to use the graphic organizer to plan the characters, plot, setting, theme, and point of view of their story. Participants had about 10 minutes to plan their narrative using the graphic organizer. After participants planned their narrative, they had about 15 minutes to draft. When they were finished, they had about 10 minutes to revise and edit. Upon completion, participants assessed their composition using a guide (Appendix C).
Session Four: Writing Attitude Survey and Interview. Session Four took place in my classroom, lasting about two hours. During the fourth session, participants once again completed the Elementary Writing Attitude Survey (Kear et al., 2000) to determine if their attitude towards writing changed after using Google Voice Typing.
Once the survey was completed, participants were interviewed. A one-on-one semi-structured interview was used for this study.
Session Five: Follow-Up. Session Five took place in my classroom, lasting 30 minutes to 1 hour. During the fifth session, participants listened to their responses to each interview question, correcting or clarifying meaning or misconceptions.
Data Collection
Case study research uses, “detailed, in-depth data collection involving multiple sources of information” (Creswell, 2013, p. 97). Data were gathered from interviews, observations, documents, and audiovisuals. Interviews helped to obtain information
that could not be observed, such as participants’ feelings and spontaneous comments.
Observations gave first hand accounts of what occurred in the natural setting.
Documents provided a way to learn more about the person being studied. The audio and video recordings helped to fill in missing information.
Participant interview. A one-on-one semi-structured interview was used for this study. Semi-structured interviews are flexible, like a conversation, and use open-ended questions. Interviews are guided by a list of questions (Appendix E), not necessarily in any particular order, with probing questions and follow-up questions to clarify meaning. The interviewer avoided letting any personal views be known to the interviewee. The initial interview took place after the participants wrote four
compositions, and lasted approximately one hour. A second interview was scheduled for approximately 30 minutes to one hour to allow participants a chance to review the transcript and clarify any points.
During the semi-structured interview, participants were asked questions over three major domains: (a) participant’s response to using voice-to-text to compose, (b) the effect of voice-to-text on the content, appearance, or other features of compositions, and (c) how voice-to-text changed the composing process for participants. From these domains, the participants were asked questions, such as,
• “What was difficult or easy about using Google Voice Typing?”
• “How would you describe your writing after using Google Voice Typing?”
• “How do you think Google Voice Typing influenced (helped or worsened) your writing?”
After each interview was completed in session four, the audio and video recordings were used to transcribe interviews. Then, the interview was edited in order for the participants to be able to read it. During the member check in session five, participants were given a chance to correct, clarify, or make changes to their answers.
After the member check, I read back through the interview and added analytical memos containing some preliminary analysis and interpretations in my journal.
Observations. Observations occurred in the setting of the phenomenon.
According to Criswell (2013), when “the researcher is participating in the activity at the site,” (p. 166) he or she is a participant as observer. Observations were documented on the observation checklist (Appendix F), with details about the physical setting, the participants, activities and interactions, conversations, and subtle factors. Additionally, participants were recorded via audio and video during the Writing Technology
Workshop in order to get a full account of what occurred. To add to the field notes that were gathered during each session, I watched and listened to the video and audio recordings for each session, totaling about eight hours per participant. While watching and listening, I added to field notes and then wrote a highly descriptive, rich, and reflective narrative. While writing the narrative, I added to analytical memos containing some preliminary analysis and interpretations in my journal.
Documents. Data were collected using documents from four sources:
Elementary Writing Attitude Surveys (Kear et al., 2000), compositions, rubrics, and narrative guides. These documents are considered researcher-generated documents because they were “…prepared by the researcher or for the researcher by participants after the study has begun” (Merriam & Tisdell, 2016, p. 174). Each participant took the
Elementary Writing Attitude Survey (Kear et al., 2000) before writing the first composition and after writing the last composition. The survey had 28 items that presented a brief, simply worded statement about writing. Following the statement were four Garfield pictures in which Garfield’s mood appears to be very happy, somewhat happy, somewhat upset, and very upset. After reading the statement,
participants’ circled the Garfield that best represented their feelings. The survey gave a quick indication of each participant’s attitude toward writing.
Writing conditions. Participants created four compositions under four different writing conditions, which included two levels of transcription (i.e., handwritten and Google Voice Typing) and two levels of planning (i.e., spontaneous writing and planned writing). The compositions were a valuable source of data as they provided a visual for participants when answering the interview questions. Additionally, the compositions were instrumental during the analysis stage.
After each composition was published, participants assessed the composition using the Narrative Mode Guide (The Culham Writing Company, 2016a) or Expository
After each composition was published, participants assessed the composition using the Narrative Mode Guide (The Culham Writing Company, 2016a) or Expository